This page contains automatically translated content.

Back
05/11/2022 | Integrated energy systems | Final theses

Master thesis: Migrating a Decision Making Framework to Python

Developing a concept to migrate and integrate existing Java decision framework to the python-based ABM-framework Mesa and according implementation

Background

LARA (Lightweight Architecture for boundedly Rational Citizen Agents) is a component-based framework to model agents’ decision making between behavioural options. It includes pre- and postprocessing of options, memory, and option selection. Originally written in Java, it now shall be applied to python models as well.

Your Profile

  • Ongoing studies in computer science with good programming skills in python and/or Java

  • Interest in agent-based modelling

  • Preference to understand, implement and test a decision making modelling component

  • Proactivity, capacity for teamwork, and good communication skills

  • Good knowledge of German and/or English (both orally and written)

You may expect

The section “Integrated Energy Systems“ explores the transition of current energy systems with three research foci: energy economics and decision support, coordination and communication, as well as wind energy.

The group Communication and Coordination in the Energy System (COOKIES) develops and applies agent-based models to analyse individual investment decisions. This enables investigations about the impact of changes in the regulatory framework and other measures on dynamics of acceptance and investments of individual actors and their interaction.

Working at the section means profiting from the diverse expertise of an engaged team and access to the Fraunhofer IEE, with which we cooperate.

Your tasks

  • Introduction to the conception and function of a decision making framework to model agents’ decisions

  • Developing a concept to migrate and integrate existing Java code to the python-based ABM-framework Mesa and according implementation

  • Testing of the implementation with existing agent-based models

  • Adaptation of existing documentation

Contact

Dr. Sascha Holzhauer